HAL Id: hal-02971510
https://hal.archives-ouvertes.fr/hal-02971510
Submitted on 21 Dec 2020
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
The necessity of investigating a freshwater-marine continuum using a mesocosm approach in nanosafety:
The case study of TiO2 MNM-based photocatalytic cement
Amélie Châtel, Melanie Auffan, Hanane Perrein-Ettajani, Lenka Brousset, Isabelle Métais, Perrine Chaurand, Mohammed Mouloud, Simon Clavaguera,
Yohann Gandolfo, Mélanie Bruneau, et al.
To cite this version:
Amélie Châtel, Melanie Auffan, Hanane Perrein-Ettajani, Lenka Brousset, Isabelle Métais, et al.. The necessity of investigating a freshwater-marine continuum using a mesocosm approach in nanosafety:
The case study of TiO2 MNM-based photocatalytic cement. NanoImpact, Elsevier, 2020, 20, pp.1-8.
�10.1016/j.impact.2020.100254�. �hal-02971510�
The necessity of investigating a freshwater-marine continuum using a mesocosm approach 1
in nanosafety: the case study of TiO2 MNM-based photocatalytic cement 2
Amélie Châtel1*, Mélanie Auffan2,3, Hanane Perrein-Ettajani1, Lenka Brousset4, Isabelle 3
Métais 1, Perrine Chaurand2, Mohammed Mouloud1, Simon Clavaguera5, Yohann Gandolfo1, 4
Mélanie Bruneau1, Armand Masion2, Alain Thiéry4, Jérôme Rose2,3, Catherine Mouneyrac1 5
1Laboratoire Mer, Molécules, Santé (MMS, EA 2160); Université Catholique de l’Ouest, Angers F-49000 France
6
2CEREGE, CNRS, Aix Marseille Univ, IRD, INRA, Coll France, Aix-en-Provence, France
7
3Duke University, Civil and Environmental Engineering, Durham, USA
8
4Aix Marseille Université, Avignon Université, CNRS, IRD, IMBE UMR 7263, FR— 13284, Marseille
9
5Université Grenoble Alpes, Commissariat à l’Energie Atomique et aux Energies Alternatives (CEA), LITEN, NanoSafety
10
Platform, F-38054 Grenoble, France
11
*amelie.chatel@uco.fr
12
Abstract 13
Production of Manufactured Nanomaterials (MNMs) has increased extensively due to 14
economic interest in the current years. However, this widespread use raises concern about 15
their impact on human and environment. Current efforts are made, both at national and 16
international levels to help developing safer MNMs in the market. In order to assess hazards 17
of MNMs, it is important to take into account exposome parameters in order to link fate and 18
behavior of MNMs to their potential toxicity. In that context, the aim of this study was to 19
compare exposure and impact of TiO2 MNMs-based cement at different levels of its life cycle 20
(TiO2MNMs, cement containing TiO2 MNMs) between two exposure mesocosm scenarios 21
mimicking : marine conditions using the bivalve Scrobicularia plana and freshwater 22
conditions using the gastropod Planorbarius corneus, for 28 days. These approaches allows 23
measurements of physical-chemical parameters throughout the duration of the exposure.
24
Similar results were observed in both exposure conditions since in the two scenarios Ti was 25
removed from the water column and accumulated in surficial sediments. While in P. corneus, 26
statistically different concentrations of Ti were measured in the digestive glands compared to 27
controls following exposure to TiO2 MNMs, elevated background of Ti concentrations were 28
measured in the controls of S. plana that did not allow to discriminating any bioaccumulation 29
process. In addition, both TiO2 MNMs and TiO2MNM-based cement exposed S. plana did not 30
present any activation of the p38 mitogen-activated protein kinase (MAPKs).This study 31
demonstrates the challenge of using freshwater-marine continuum using a mesocosm 32
approach in nanosafety.
33
Keywords:Mesocosm, TiO2MNMs, Scrobicularia plana, Planorbarius corneus 34
INTRODUCTION 35
Releases of Manufactured Nanomaterial (MNMs) can occur during any stage of the 36
MNM life cycle i.e. production stage of nanoproduct, the use/application stage of these 37
nanoproducts and their ultimate end-of-life management/disposal. A comprehensive 38
understanding of the potential for such releases along the whole life cycle and their possible 39
effects is crucial to ensure the safe and sustainable use of these new materials (Salieri et al., 40
2018). Among all MNMs, TiO2 MNMs are introduced into many products such as paints, 41
plastics, food additives, sunscreens and other care products (Yin 2012). The use of TiO2
42
MNMs into cement is a recent application in building material industry. Thanks to its 43
photocatalytic properties, it confers to the building material air decontamination, self- 44
sterilizing, self-cleaning and anti-fogging abilities (Jimenez et al., 2016). For those reasons, 45
photocatalytic cement appears to be an appealing market for industry that represents from 0.1 46
to 1% of the total European production of TiO2 and hence about 10.2t to 102t of TiO2 MNMs 47
released each year (Nanotechproject.org). TiO2MNMs incorporated in the cement matrix are 48
usually non-coated anatase with a mass concentration ranging from 0.3 to 10 wt% (Ruot et al., 49
2009). Degradation of TiO2 MNM-based cement is likely to happen at each stage of the 50
cement life cycle (production, manufacturing, use disposal and recycling) (Bossa et al., 2017).
51
The main causes of release of TiO2MNMsfrom photocatalytic coatings is the stripping of the 52
TiO2 from the coatings because of the water flow, the dispersion of agglomerate under NaCl 53
and UV light condition and finally its discharge from loosening caused by mechanical damage 54
(Olabarietta et al., 2012). A recent study showed that TiO2 MNMs were released (18.7 ± 2.1 55
to 33.5 ± 5.1 mg of Ti/m2 of cement after 168 h) from photocatalytic cement pastes that were 56
leached at a lab-scale to produce a range of degradation rates, meaning that in the worst-case 57
scenario of weathering a negligible mass of TiO2 (0.04w.%) was released from the 58
photocatalytic cement (Bossa et al., 2017). TiO2 MNMs release comes from a very thin active 59
surface layer where both the cement surface chemistry and its pore network morphology 60
control the TiO2MNMs diffusion (Bossa, 2019).
61
Laboratory studies have extensively demonstrated the adverse effects of TiO2 MNMs 62
on many aquatic organisms (Ramsden et al., 2009; Binh et al., 2016; Hall et al., 2009; Kulacki 63
et al., 2012). Nevertheless, those studies have been conducted in simple systems (microcosm) 64
and they did not reflect the MNMs fate and behavior in the environment due to processes such 65
as sorption and aggregation of MNMs (Auffan et al., 2014). As a consequence, for a strongly 66
characterization of risk, the combination of the organism, its environment and MNMs needs 67
to be taken into account and requires interdisciplinary expertise in physical-chemistry, 68
biology and ecotoxicology (Auffan et al., 2014). In that context, mesocosms represent 69
relevant experimental systems for investigating the complex issue of exposure that enable to 70
get quantitative time- and spatial data on the distribution of MNMs within a simulated 71
ecosystem. A mesocosm refers to “an experimental system that simulates real-life conditions 72
as closely as possible, while allowing the manipulation of environmental factors” (FAO, 73
2009). As stated during the H2020 European commission funded NANoREG project (Project 74
#646221), mesocosms have been shown to provide a reliable methodology to obtain 75
quantitative time- and spatially regarding exposure-driven environmental risk assessment.
76
In this context, two indoor mesocosm platforms (one marine and one freshwater) were 77
conceived with small size tanks (60 L) for investigating MNM exposure and impacts on 78
aquatic species that allows to recreate freshwater and estuarine conditions (sediment, tide 79
cycles, controlled temperature, salinity and redox potential) (Auffan et al. 2019). TiO2 MNMs 80
and TiO2 MNM-based cement fate, behavior, bioaccumulation and toxicity were evaluated 81
after a 28 day exposure in two mollusks representative of the estuarine and freshwater 82
ecosystems : the bivalve Scrobicularia plana and the gastropod Planorbarius corneus.
83
Scrobicularia plana is an endobenthic bivalve involved in the functioning and the structure of 84
estuarine ecosystems and has also been largely demonstrated to represent a relevant model for 85
biomonitoring (Mouneyrac et al., 2014). Planorbarius corneus is a hermaphroditic snail that 86
inhabits small temporary ponds. This species belongs to the Planorbidae, the largest family of 87
aquatic pulmonate gastropods distributed all over the world (Jopp, 2006). Planorbarius 88
corneus play an important role in trophic chains as grazer and as prey (Wojdak and Trexler, 89
2010) and have already been used as model organisms for exposure and toxicity studies in 90
mesocosms (Tella et al 2014, 2015, Auffan 2018).
91
The originality of the project was to compare the fate and behavior of a nano-enable product 92
(TiO2 MNM-based cement) in these two freshwater and marine ecosystems. Two stages of the 93
MNM lifecycle were considered : the formulation stage using bare TiO2MNMs and end of life 94
stage of the cement containing TiO2MNMs using cement leachate. Freshwater and marine 95
mesocosms were used to obtain resolved data on the distribution, transformation and impact 96
towards these mimicked ecosystems depending on the physical-chemical properties of the 97
environment (salinity, pH, conductivity, tide cycle or not…) and the physical-chemical 98
properties of the contaminant at two stages of the lifecycle. One of the major drawbacks in 99
assessing environmental impact of MNMs is that the environmental concentrations are not yet 100
known. Herein, we used between 1and 1.2 mg.L-1of TiO2 MNMs which belong to the 101
threshold values estimated in highly exposed areas (Bourgeault et al., 2015).
102
MATERIAL AND METHODS 103
TiO2NPs 104
Primary size and shape of TiO2 MNMs (NM212, JRC repository) were determined by an 105
ultra-high resolution scanning electron microscopy SEM LEO 1530 (LEO Electron 106
Microscopy Ltd, Cambridge, England) and Transmission Electron Microscopy (analytical 107
TEM Field Emission Gun (FEG) 200 KV Osiris from Tecnai, FEI, Japan). The SEM and 108
TEM images corroborate the nanometric size of the particles (see SI, figure S1). X-ray 109
diffraction (XRD) patterns of the samples were obtained using a Bruker D8 X-ray diffraction 110
(XRD) system in a θ-2θ mode and Cu Kα X-ray source. The TiO2 MNMs have a body- 111
centered tetragonal anatase crystal structure (see SI). The width of the XRD peaks were used 112
to calculate a crystallite size about 9 nm. A specific surface area of 223 m2/g was determined 113
by the Brunauer-Emmett-Teller (BET) adsorption method using the N2 adsorption isotherm 114
measured at 77K (BELSORP-max, BEL Japan Inc.).
115 116
TiO2 MNMs-based cement 117
Photocatalytic white Portland cement incorporating TiO2 MNMs was provided by a European 118
cement manufacturer (Calcia, France) through the French Technical Association of Industries 119
of Hydraulic Binders (ATILH, France). This cement is mainly composed of Ca and Si (≈66wt 120
% CaO and ≈23 wt% SiO2) and contains 2.85 wt% of TiO2 (anatase) (Bossa et al., 2017).
121
Cylindersof hardened cement (4 cm diameter, 8 cm high) were obtained after 28 days of 122
hydration at 20°C with a water/cement ratio of 0.5. Hardened cement pastes exhibit a complex 123
mineralogy, i.e. a mixture of hydrated minerals (portlandite, calcium silicate hydrates, 124
ettringite) and residual anhydrous minerals (di and tricalcium silicates). TiO2MNMs are 125
homogeneously dispersed in the hydrated paste, with the exception of few high spotsof50 µm 126
(Bossa et al., 2017).
127
Cement degradation residue generation was simulated using accelerated lab-scale approach to 128
conduct a so-called « worst-case scenario » and to generate a quantity of cement degradation 129
residues for mesocosms dosing. Finely crushed cement pastes were leached in batch using 130
ultra-pure water (UPW) with an elevated liquid-to-solid ratio (L/S) of 100. After pH 131
stabilization (pH > 12.5), the suspension of cement degradation residues was neutralized with 132
slow addition of nitric acid (22.37 mol.L-1) reach a pH of 7.8 and a conductivity of 13 mS.cm- 133
1. The solution is composed of a dissolved fraction (mainly Ca and Si, 203 and 3.96 mg.L-1 134
respectively) and a particulate fraction with a TiO2 concentration of 143.7±2.6 mg.L-1. In 135
accordance with the very low TiO2 MNM solubility, no TiO2 MNM dissolution was observed 136
(Table S1). The concentrations of Ca and Si in the dissolved fraction reveal the high 137
degradability of the cement matrix in natural media (pH 7.8). Presence of free TiO2 MNMs, 138
i.e. not embedded in cement matrix, in the solution of cement degradation residues is then 139
suspected. In neutralized cement degradation residues, TiO2 MNMs were observed mainly 140
associated with aggregates larger than 0.45 µm and with SiAl intermix chains. SiAl chains are 141
supposed to be geopolymers stable at pH 7 and resulting from secondary precipitation/
142
polymerization during the pH neutralization process (Bossa et al., 2017).
143 144
Animals 145
Individuals of Scrobicularia plana (15-20 mm length) were hand-collected in November 2015 146
in the intertidal mudflat on the French Atlantic coast (Bay of Bourgneuf: N 47°01’50.35”, W 147
1°59’04’80’’) and transported in cool boxes to the laboratory. Then, bivalves were 148
immediately placed into aerated artificial water (Tropic marin®) at 30 practical salinity units 149
(psu) during 5 days in a temperature controlled room at 15°C (temperature experienced in the 150
field at the collection time). A natural inoculum containing picoplankton as primary producer 151
(bacteria, algae, protozoa, etc.) was also collected in the intertidal mudflat. Physico-chemical 152
parameters of seawater from the collection site were : pH= 7.7; salinity 28 psu; conductivity 153
40 mS.cm-1. 154
Individuals of Planorbarius corneus (L., 1758) (Great Ram’s Horm snail, benthic grazer) 155
were hand-collected in October 2015 in a non-contaminated temporary pond part of the 156
Natura 2000 reserve network in south of France (Bonne Cougnes: N 43°20’47.04”, E 157
6°15’34.786’’, 246 m a.s.l.) and transported in pond water to the laboratory. A natural 158
inoculum containing picobenthosas primary producer (bacteria, algae, protozoa, etc.) was also 159
sampled in the pond. Physico-chemical parameters of the pond water from the collection site 160
were: water temperature 14.9°C; conductivity 819 mS.cm-1; concentration of dissolved 161
oxygen 0.72 mg.L-1 (6.6%).
162 163
Mesocosm setup 164
Freshwater mesocosms. The mesocosm experiments were already described by Auffan et al.
165
(2014), Tella et al. (2014), Tella et al. (2015), Auffan et al. (2018)(figure 1). Tanks (750 x 200 166
x 600 mm) were filled with 6-8 cm artificial sediments containing 89% SiO2, 10% kaolinite 167
and 1% of CaCO3 w/w adapted from OECD guideline (OECD, 2004). Primary producers 168
were brought by ~300 g of water-saturated natural sediment (sieved at 200 µm) laid at the 169
surface. The mesocosms were then gently filled with 50 L of Volvic water with pH and 170
conductivity close to the natural pond water (pH 7, 11.5 mg.L-1 Ca2+, 13.5 mg.L-1 Cl-, 71 171
mg.L-1 HCO32-, 8 mg.L-1 Mg2+, 6.3 mg.L-1 NO3-, 6.2 mg.L-1 K+, 11.6 mg.L-1 Na+).
172
After 2 days, the physical-chemical parameters of the mesocosms were stabilized (turbidity, 173
pH, dissolved O2, redox). Then, 19 adult P. corneus were added per mesocosm and 174
acclimatized for one week. Organisms were exposed for 28 days in a temperature controlled 175
room (18°C) and under controlled light (photoperiod day/night 10:14). Physico-chemical 176
parameters including temperature (°C), pH, redox potential in the water column and the 177
sediment (mV), and dissolved oxygen (mg O2.L-1) were monitored every 2 min during the 178
whole duration of the experiment. Among the 6 freshwater mesocosms, 2 were kept as 179
negative control (without contamination), 2 were exposed to TiO2 MNMs (considered as 180
positive control), and 2 were exposed to cement leachate. While the first control (negative 181
control) condition allows to follow animal behavior (mortality) during the experiment, the 182
second one serve as a positive control experiment to follow TiO2 bioaccumulation by the 183
organisms in order to compare with animals treated with cement for both freshwater and 184
marine systems.
185 . 186
Marine mesocosms. Each experimental unit was composed with two tanks (mesocosm and 187
reserve tank), two pumps (Eheim compact 300L.h-1) and a mechanical timer (IDK PMTF 188
16A) allowing to mimicking the tidal cycle (figure 1). The mesocosm tank (750 x 200 x 600 189
mm) was filled with 14 kg synthetic sediment (90% SiO2, 9% kaolinite, 1% CaCO3 w/w) 190
adapted from OECD guideline (OECD, 2004). Three hundred g of water-saturated natural 191
inoculums was laid at the surface to bring primary producers. Then, each mesocosm tank was 192
flooded with 35 L of the artificial seawater Tropic Marin™ (Tropicarium Buchshlag Dreieich, 193
Germany) at 30psu. The reserve tank (700 x 500 x 37 mm) was placed under the first one 194
(mesocosm tank) to collect removed water from the mesocosm tank during low tide.
195
Reservoir tank was always filled with a minimum of 25 L of artificial seawater at 30 psu to 196
constantly maintain the measuring probe immersed.
197
After one week, the physical-chemical parameters of the mesocosms were stabilized 198
(turbidity, pH, dissolved O2, redox). Then, 55 S. plana were added per mesocosm and 199
acclimatized for 5 days. Organisms were exposed for 28 days in a temperature controlled 200
room (15°C), under controlled light (photoperiod 16:8) and at a tidal cycle (6h of low tide, 6h 201
of high tide; two tides /day). Physical-chemical parameters including temperature (°C), 202
salinity (psu), pH, redox potential (mV), dissolved oxygen (mg O2.L-1) were monitored every 203
15 min into the water column from the reserve tank during the whole duration of the 204
experiment. Among the 9 marine mesocosms, 3 were kept as negative control (without 205
contamination), 3 were exposed to TiO2 MNMs (positive control) and 3 were exposed to 206
cement leachate.
207
208
Figure1. Scheme of the freshwater (left) and marine (right) mesocosms.
209 210
Mesocosm dosing 211
Aqueous suspensions of TiO2 MNMs and TiO2 MNM-based cement degradation residues 212
(with a stabilized pH of 7.8) were prepared prior injections in the mesocosms. Aliquots of 213
these suspensions were digested using an Ultra WAVE microwave digestion system with 1 214
mL HNO3 67% (NORMATOM) and 1 mL HF 47%-51% (PlasmaPure) and analysed by ICP- 215
MS (Perkin Elmer© Nexion 300 ICP-MS) for their total Ti contents. A concentration of 216
4.86±0.13 g.L-1 TiO2(n=3) was found for the TiO2 MNM suspension (2.91±0.08 g.L-1Ti). The 217
TiO2 concentration measured in the solutions of cement degradation residues (n=6) was 218
143.7±2.6 mg.L-1 (86±1.6 mg.L-1Ti). Both suspensions were close to the targeted 219
concentrations and showed excellent recovery yields after digestion (> 95%).
220
A multiple dosing experiment was performed on a 4 week-period. A total of 12 dosings were 221
achieved (3 times per week), corresponding to 0.09 mg.L-1 of TiO2 per injection in order to 222
reach a final nominal concentration of 1mg.L-1 TiO2 per mesocosm. As the suspension of 223
cement degradation residues is not stable over time, a new neutralized batch was prepared 224
every two injections and stored at 4°C in dark between the two injections.
225
Odeon
Multi parameters probes
Data logger ORP probes
Water flow Light
Pu mp
Pu mp
Air diffus er
The real concentration injected over time was determined by ICP-MS and are plotted in 226
Figure 2. For each injection, a low and realistic concentration of 0.1 mg.L-1 of injected TiO2 227
was targeted in the mesocosm (0.06 mg.L-1Ti). After 30 days (12 injections), the cumulated 228
concentration of TiO2 injected reached 1-1.2 mg.L-1 in the mesocosms.
229 230 231
232
Figure 2. TiO2 injection sequence for TiO2MNMs and TiO2MNM-based cement degradation 233
residues in both freshwater and marine mesocosms over the 28 days-contamination period.
234 235
Ti quantification in the different compartment of the mesocosms 236
The concentration of Ti in the mesocosms following TiO2 MNMs or cement leachate 237
contamination was measured in the water column, in the surficial sediments and in the 238
organisms sampled at 0, 7, 14, 21 and 28 days. Water was sampled at ~10cm from the 239
air/water interface. Surficial sediments (between 500 µm to 1000 µm depth) were sampled at 240
three different locations and then pooled before being dried. Samples were digested using 241
microwaves at 180°C before analysis by ICP-MS.
242
For the marine mesocosms, Ti quantifications were performed by Micropolluants Technologie 243
S.A. (Metz, France) using an Agilent 7700 ICP-MS (Agilent Technologies France). 10 mL of 244
water or 2 mL of surficial sediments were mixed with 1mL 32-35% HCl, 1mL 67-69% HNO3, 245
100 µL 94-98% H2SO4, and 1 mL 47-51% HF prior ICP-MS measurements. For S. plana, 246
whole soft tissues were mixed with 1 mL 67-69% HNO3, 100 µL 94-98% H2SO4, and 1 mL 247
47-51% HF. For the freshwater mesocosms, water samples (2 mL) were digested with 1 mL 248
HNO3 67-69% and 0.5 mL HF 47%-51%. For the surficial sediments (50 mg), a mixture of 3 249
acids (1 mL HCl 34%, 2 mL HNO3 67%, 0.5 mL HF 47%-51%) was used. For P. corneus, the 250
digestive glands were dissected, before being digesting using 1 mL HNO3 67%, 0.5 mL H2O2 251
30%-32% and 1 mL HF 47%-51%. Samples were digested using the microwave Ultra WAVE 252
and the analysis were performed using the Perkin Elmer Nexion 300 ICP-MS. Using 253
sediments and freshwater organisms spiked with TiO2 nanoparticles, a recovery yield of 254
>95% after microwave digestion and ICP-MS analysis was obtained.
255 256
Evaluation of immunotoxic effects in S. plana: pp38 level measurement by Western blot 257
Total whole soft tissues of S. plana exposed for 28 days (n=7 for each condition) were 258
homogenized with lysis-buffer (NaCl 0.2M, DTT 1mM, protease inhibitor cocktail 0.1%) in a 259
mortar at 4°C. The clear supernatant was then obtained by centrifugation (13,000xg, 10 min, 260
4°C) and stored at -20°C until analysis. Protein concentrations were determined according to 261
Bradford (1976) using bovine serum albumin as a standard. Protein extracts were mixed with 262
sample buffer (0.5 M Tris–HCl, pH 6.8, 2% SDS, 10% glycerol, 4% 2-mercaptoethanol, 263
0.05% bromophenol blue) 1:3 and boiled for 5 min at 95°C. Proteins (40 µg per lane) were 264
separated using a 12% polyacrylamide gel. Western blot analysis was performed as previously 265
described (Châtel et al., 2011b). After transfer, membranes were incubated with a mouse 266
monoclonal antibody anti-Phospho-p38 MAPK (Thr180/Tyr182) (9216-Cell signaling 267
Biotechnology) diluted in 1/1000 in TBS/BSA 5% or with a mouse monoclonal antibody anti- 268
β-Actin (A5441 -SIGMA) diluted in 1/2000 in TBS/BSA 5%. After washing (3 times for 5 269
min), membranes were incubated with a goat anti-Mouse IgG antibody coupled with alkaline 270
phosphatase (A3562, Sigma Aldrich) diluted 1:2000 in TBS/BSA 5%. Detection of immune 271
complexes was carried out by colorimetric reaction using a solution of 5-bromo-4-chloro-3- 272
indolyl phosphatase (BCIP) and nitroblue tetrazolium (NBT). After visualization, membranes 273
were scanned and band intensity, which is the relative amount of protein expression, was 274
quantified as Relative Optical Densities (ROD; pixels/mm2). Results of pp38 were normalized 275
according to actin expression.
276 277
Statistical analysis 278
The measured values were compared among different groups using the non-parametric test 279
Mann–Whitney (XL-Stat software) and the Student’s t-test. Statistical significance was 280
accepted at p<0.05.
281 282
283 284
RESULTS and DISCUSSION 285
286
Favorable physical-chemical conditions in freshwater and estuarine mesocosms 287
During the 28 days of contamination several physical-chemical parameters were monitored to 288
assess the global response of the mesocosms to the presence of TiO2 MNMs and cement 289
leachate. In marine as well as in freshwater mesocosms, the average temperature, dissolved 290
O2, conductivity, and pH were not significantly different between negative controls and 291
contaminated mesocosms(see SI, Figures S2 and S3).
292
Oxido-reductive probes indicated that freshwater was oxidative (310±20 mV), while reductive 293
conditions prevailed in the sediments (ca -370 mV). Regarding the conductivity, a step by 294
step increase from day 0 to day 28 (200 ± 1 µS.cm-1 to 288 ± 0.4 µS.cm-1respectively) was 295
observed. Conductivity drops were recorded during the weekly refills with ultrapure water to 296
compensate the evaporation. On the whole, no significant differences were observed between 297
controls and contaminated mesocosms.
298
Consequently, during the exposure to TiO2 MNMs and cement leachate, the physico-chemical 299
conditions of the 15 mesocosms remained favorable with the oxygenation, pH, temperature, 300
redox potential in the range of natural conditions.
301 302
Homo-aggregationof TiO2MNMs and accumulation in the surficial sediments 303
Bossa et al. (2017) estimated using an accelerated aging protocol that a negligible mass of 304
TiO2 (0.04w.%) was released from the photocatalytic cement (Bossa et al., 2017). Herein, we 305
used an accelerated lab-scale approach to conduct a so-called « worst-case scenario » 306
assuming that 100% of the Ti contained in the cement would have been released. Figure 3 and 307
4 show the Ti concentrations measured in the water column and sediments of both freshwater 308
and marine mesocosms as a function of time. Ti was detected in both waters and sediments 309
sampled in control mesocosms proving that a high geochemical background of this element 310
existed in the mimicked ecosystems. Average background Ti concentrations in freshwater 311
mesocosms were 19.2±17.2 µg.L-1in the water column and 2230±928 mg.kg-1 in surficial 312
sediments. Average background Ti concentrations in marine control mesocosms were 313
567±273mg.kg-1 in surficial sediments and below the quantification limit (<2.5 µg.L-1) in the 314
water column. Ti is a naturally occurring element in mineral and amorphous phases occurring 315
in freshwater and marine environments. Ti can be found in titanium-iron oxide minerals (as 316
ilmenite), TiO2 minerals as rutile, anatase, and brookite. Herein, the TiO2 MNMs used were 317
engineered nanosized anatase minerals. Consequently even if the Ti background concentration 318
was elevated, it is noteworthy that the speciation and reactivity of the naturally occurring Ti- 319
bearing phases might be different from the MNM injected.
320
In the surficial sediments of marine mesocosms, high concentrations of Ti were detected after 321
7 days (control, TiO2 MNMs, and cement leachate) and progressively decreased until 28 days 322
(Figure 4, right). However, because of the elevated geochemical background of Ti, no 323
statistical difference was observed between controls and contaminated mesocosms. Similarly, 324
no increase in Ti concentration could be evidenced in the surficial sediments of contaminated 325
freshwater mesocosms compared to controls. Besides, the fluctuations of Ti concentration 326
measured over time were likely related to the heterogeneity of the Ti distribution and the 327
difficulty of sampling the surficial sediments(Figure 4). These results highlighted the 328
challenge of a reliable quantification of Ti in that compartment.
329 330
Total Ti concentrations in the water column of marine mesocosms (control, TiO2 MNMs and 331
cement leachate) were always below the quantification limit estimated at 2.5 µg.L-1 whatever 332
the duration of exposure. On the opposite, total Ti concentrations in the water column were 333
always above 10 µg.L-1 in freshwater mesocosms. The only condition for which a statistically 334
significant difference was observed between the water column of controls and contaminated 335
freshwater mesocosms was after 7 days of exposure to TiO2 MNMs. At this time point, the 336
difference between TiO2 MNMs contaminated mesocosms and geochemical Ti background 337
were 2.2±1.2 µg.L-1Ti for a total Ti concentration injected of 180 µg.L-1Ti (300 µg.L-1 of 338
TiO2). Based on these values, we estimated that 98-99 % of the Ti injected was removed from 339
the water column after 7 days. Although no increase in Ti concentration could be evidenced in 340
the surficial sediment of freshwater mesocosms, we hypothesize that both cement leachate 341
and TiO2 MNMs settled at more than 98% at the surface of sediments. Based on the 31.0±0.6 342
mg and 35.0±0.9 mg Ti introduced respectively after 28 days in mesocosms contaminated 343
with cement leachate and TiO2 MNMs, the remaining of 2% of Ti particles in the water 344
column should entail a theoretical increase in the surficial sediment of176 mg.kg-1, which is 345
far below the standard deviation determined for Ti background in that compartment. This 346
explains why Ti accumulation at sediment surface remained undetected by chemical analysis.
347
In addition, in marine mesocosms, due to the tidal system, part of the sediment were removed 348
along with the water in the lower tank and sedimentation of TiO2 occurred as well in this 349
compartment favoring decrease in its concentration as compared to freshwater mesocoms.
350 351
352
Figure 3. Total Ti concentration in the water column after 7, 14, 21 and 28 days of exposure 353
in freshwater mesocosms. Ti concentration in the water column in marine mesocosms is not 354
represented as values were always below the quantification limit estimatedat 2.5 µg.L-1. 355
Statistical significance was determined by the Student’s t-test (*: p<0.05).
356 357
358
Figure 4. Total Ti concentration in surficial sediment after 7, 14, 21 and 28 days of exposure 359
in freshwater (left) and marine (right) mesocosms.
360 361
In freshwater mesocosms, the size distribution of the suspended matter in the 0.1 µm – 1000 362
µm size range and the total number of suspended particles in the 0.4 µm – 0.9 µm size range 363
were not different between control and contaminated mesocosms (Figure 5). This highlighted 364
that the main mechanism of aggregation and settling down expected in the water column was 365
the homo-aggregation of the TiO2 MNMs either pristine or brought via cement leachate. In 366
marine mesocosms, MNM homo-aggregation was expected to be even faster due to higher 367
alkalinity and ionic strength as compared to freshwater environments (Peralta-Videa et al., 368
2011; Rocha et al., 2015;Vale et al., 2016). It is noteworthy that the aggregation of MNMs 369
allows their deposition onto the surficial sediment (Peralta-Videa et al., 2011), resulting in 370
high exposure of benthic organisms compared to planktonic organisms (Mohd Omar et al., 371
2014; Mouneyrac et al., 2014).
372
*
373
Figure 5. (left) Size distribution of colloidal particles in the water column of freshwater 374
control mesocosms at day 0. The dotted lines mark out the 0.4 µm – 0.9 µm fraction. (right) 375
Evolution of the number of colloidal particles in the 0.4µm – 0.9 µm range from day 0 to 28, 376
under the three conditions tested (control, cement leachate and TiO2 MNMs) in freshwater 377
mesocosms.
378 379
Bioaccumulation and effects on benthic organisms 380
P. corneus and S. plana are two benthic grazer species. While P. corneus eat algae and 381
biofilms at the sediment/water interface(Jones 1961), S. plana can either feed suspended 382
particles from the water column but also ingest sedimentary particles (Hughes, 1969). For S.
383
plana exposed in marine mesocosms, elevated background of Ti concentrations were 384
measured in the control group (84.8±42.3 mg.kg-1 dry weight) which did not allow to 385
discriminating any bioaccumulation process in organisms exposed to TiO2 MNMs and cement 386
leachate (97.1±43.9 and 107.5±48.2 mg.kg-1 dry weight, respectively) (figure 6).
387
Background concentrations of Ti measured in the digestive glands of unexposed P. corneus 388
were one order of magnitude lower. Following exposure to cement leachate, no difference in 389
Ti concentration in the digestive glands were measured with respect to control P. corneus.
390
However, statistically different concentrations of Ti were measured in the digestive glands 391
compared to controls following exposure to TiO2 MNMs i.e.1.5 ± 0.2 mg Ti.kg-1 (dry weight) 392
and 2.5 ± 0.1 mg Ti.kg-1 (dry weight) after 21 and 28 days of exposure respectively (figure 6).
393
This suggests that an ingestion of Ti by P. corneus occurred likely resulting from the 394
accumulation of TiO2 MNMs on the surficial sediments.
395
Some authors have shown abilities to discriminate natural Ti from TiO2 NMs. For example, 396
Bourgeault et al. (2015) resolved this lack of sensitivity using isotopically modified TiO2 397
nanoparticles (47Ti) to characterize the processes governing Ti bioaccumulation in a 398
freshwater environment. Thanks to the 47Ti labeling, they detected bioaccumulation of NPs in 399
Dreissena polymorpha exposed for 1 h at environmental concentrations via water (7–120 400
μg/L of 47TiO2 NPs) and via their food (4–830 μg/L of 47TiO2 NPs mixed with 1 × 106 401
cells/mL of cyanobacteria) despite the high natural Ti background, which varied in individual 402
mussels. Such a methodology would be particularly relevant in mesocosms experiment 403
mimicking real environments with high Ti background.
404 405 406
407
Figure 6. Total Ti concentration in the digestive gland of P. corneus (left) and total soft 408
tissues of S. plana (right) after 7, 14, 21 and 28 days of exposure in mesocosms. The 409
statistical significant was determined by the Student’s t-test (*: p<0.05; ** : p<0.01).
410 411
During the 28 days of exposure in mesocosms, the survival rates of P. corneus and S. plana 412
(<2% of mortality, data not shown), picoplankton and picobenthos (see SI, Figure S4) were 413
not affected by the presence of TiO2 MNMs and cement leachate. Hence, no acute toxicity 414
was observed towards these two trophic levels. Nevertheless, a thorough characterization of 415
the biological responses (at the individual, sub-individual, and community levels) is still 416
needed to better understand the biological mechanisms of interactions.
417
One of the main effect associated with TiO2 toxicity was the induction of immune responses.
418
The MAPKs represent a superfamily of protein Ser/Thr-kinases, highly conserved through 419
evolution (Kyriakis and Avruch, 2001; Roux and Blenis, 2004), that can transduce stress 420
signals into cellular responses (Kultz and Avila, 2001; Kyriakis and Avruch, 2001). Three 421
subfamilies of the MAPKs have been well described in mammals: the extracellular regulated 422
protein kinase (ERK), the c-Jun NH2-terminal kinases (JNK) and the p38-MAPK. ERKs have 423
been more associated to cell division, growth and differentiation whereas JNKs and p38- 424
MAPKs are activated in various types of stress going from osmotic stress to chemical stress 425
and can trigger cell survival or death (apoptosis), depending on their isoform and/or cell type 426
(Canesi et al.,2001, 2002, 2006; Châtel et al., 2009, 2011a,b). They have been demonstrated 427
to play a key role in immune responses in mammals (Caffrey et al., 1999) but also in 428
invertebrates and more particular in bivalves (Canesi et al., 2001; 2002; 2006). Moreover p38 429
MAPK have been shown to be induced by TiO2 exposure in bivalves (Couleau et al., 2012).
430
P38 MAPK in S. plana have been studied in the laboratory and antibody matching has already 431
been verified on this specie as opposed to P.corneus. Since no ingestion of Ti by S. plana was 432
identified by ICP-MS, western blot analysis of S. plana using specific antibodies directed 433
against the phosphorylated form of p38 MAPK were used (Figure 7).
434
In the present study, membrane revelation showed that the molecular weight of pp38 in S.
435
plana is around 38 KDa. After normalization of the different bands representing pp38 436
expression with actin, results revealed that there was no significant modification in pp38 437
expression level for any of the conditions tested (TiO2 MNMs or cement leachate), as 438
compared to control group (figure 7). This suggested no stimulation of immune response of S.
439
plana following 28 days of exposure. Couleau et al. (2012) observed in the freshwater mussel 440
D. polymorpha exposed to TiO2 (0.1 to 25 mg.L-1 during 24 h), that ERK ½ was activated for 441
all concentrations tested in this study whereas p38 activation was only observed when 442
bivalves were exposed to 5 and 25 mg.L-1) indicating that p38 phosphorylation was less 443
sensitive than ERK ½ to this contaminant. This could be the case in the present study.
444
Moreover, an in vitro M. galloprovincialis hemocyte exposure to TiO2 (10 µg.ml-1) for 5 to 60 445
min showed that pp38 activation was transient, with an increase in phosphorylated form of 446
p38 after 5 and 15 min, followed by a dephosphorylation at 30 and 60 min (Canesi et al., 447
2010). Even though p38 is implicated in immune responses, absence of p38 activation does 448
not mean that immune response is not present. However, in the present study, the long 449
duration of exposure (28 days) could explain the absence of pp38 response depicted in S.
450
plana and investigation of its response at shorter time points would have been interesting to 451
perfom.
452
453
Figure 7. Expression of pp38 in S. plana control, TiO2 MNMs or cement leachate, analyzed 454
by Western blot. Results have been normalized to actin expression levels.
455 456
Conclusion 457
In the present study, impact of TiO2 MNMs and TiO2 MNM-based cement leachate was 458
investigated in marine and freshwater mesocosms, that allows to reproduce realistic 459
environmental exposure (mid-term duration, low doses, chronic contamination,…) and in 460
particular, a tidal system in the marine mesocosm. Such tools have been demonstrated as 461
being relevant since they take into account both hazard and exposure and hence the synergetic 462
and/or antagonistic effects of different physico-chemical parameters (temperature, pH, redox 463
potential, conductivity), that is of great interest for the regulation of the use of NMs (Auffan 464
et al., 2019).The originality of the project was also to compare the exposure and potential 465
impact of TiO2 and cement containing TiO2 MNMs on two species, representative of the 466
estuarine/freshwater continuum(with a salinity gradient),that are particular targets towards 467
pp38 Actin
38 KDa 45 KDa C C TiO2 Cement
t28days t0
0 20 40 60 80 100 120
Relative quantity of pp38 (compared to internal control actin)
C C TiO2 Cement t28days
t0
nanomaterials at the water/sediment interface. The use of similar design for both freshwater 468
and marine systems showed challenging points and limitations, that enable consistent 469
comparison of fate and behavior of NMs in different species. However, even with same 470
procedures, characterization of TiO2 in different abiotic compartments was difficult, as stated 471
in this study, and made challenging investigations of differences in bioaccumulation and 472
effects in these benthic species. The standardization of stable test suspensions for 473
nanomaterials is an ongoing challenge as it is well documented that the preparation method 474
influences the interpretation of toxicity according to the exposure. In the context of 475
developing a regulatory framework this is a challenge as it necessary that all nanomaterials 476
have equivalent preparation techniques in order to have a better standardization of 477
experiments and hence a more comparable set of data generated for nanomaterial-related data 478
categorization.
479
Finally, studying the impact of TiO2 at different stages of the TiO2 MNMs life cycle allows 480
identifying critical stages at which Safe by design (SbD) concepts can be implemented to 481
reduce adverse impacts or decrease exposure to MNMs.
482 483 484
REFERENCES 485
Auffan, M., Tella, M., Santaella, C., Brousset, L., Paillès, C., Barakat, M., Espinasse, B., 486
Artells, E., Issartel, J., Masion, A., Rose, J., Wiesner, M.R., Achouak, W., Thiéry, A. and 487
Bottero, J.Y., 2014. An adaptable mesocosm platform for performing integrated assessments 488
of nanomaterial risk in complex environmental systems. Sci. Rep. 4, 5608.
489
https://doi.org/10.1038/srep05608 490
491
Auffan, M., Masion, A., Mouneyrac, C., de Garidel-Thoron, C., Hendren, C., Thiery, A., 492
Santaella, C., Giamberini, L., Bottero, J.Y., Wiesner, M. and Rose, J., 2019. Contribution of 493
mesocosm testing to a single-step and exposure-driven environmental risk assessment of 494
engineered nanomaterials. NanoImpact. 13, 66-69.
495
https://doi.org/10.1016/j.impact.2018.12.005 496
497
Auffan, M., Liu, W., Brousset, L., Scifo, L., Pariat, A., Sanles, M., Chaurand, P., Angeletti, 498
B., Thiery, A., Masion, A. and Rose, J., 2018. Environmental Exposure of a Simulated Pond 499
Ecosystem to Cuo Nanoparticle Based-Wood Stain Throughout Its Life Cycle. Environ. Sc.:
500
Nano. 5, 2579-2589. https://doi.org/10.1039/C8EN00712H 501
502
Bour, A., Mouchet, F., Cadarsi, S., Silvestre, J., Verneuil, L., Baqué, D., Chauvet, E., 503
Bonzom, J. M., Pagnout, C., Clivot, H., Fourquaux, I., Tella, M., Auffan, M., Gauthier, L.
504
and Pinelli, E., 2016. Nanoparticles on a freshwater experimental trophic chain: A study in 505
environmentally relevant conditions through the use of mesocosms. Nanotoxicology. 10 (2), 506
245-55. https://doi.org/10.3109/17435390.2015.1053422 507
508
Bourgeault, A., Cousin, C., Geertsen, V. , Cassier-Chauvat, Chauvat, C., F., Duruphy, O., 509
Chanéac, C. and Spalla, O., 2015. The challenge of studying TiO2 nanoparticle 510
bioaccumulation at environmental concentrations:crucial use of a stable isotope tracer.
511
Environ. Sci. Technol. 49(4), 2451-9. https://doi.org/10.1021/es504638f 512
513
Caffrey, D.R., O’Neill, L.A.J. and Shields, D.C., 1999. The evolution of the MAP kinase 514
pathways: coduplication of interacting proteins leads to new signaling cascades. J. Mol. Evol.
515
49, 576- 582. http://dx.doi.org/10.1007/PL00006578 516
Canesi, L., Betti, B., Ciacci, M. and Gallo, G., 2001. Insulin-like effect of zinc in Mytilus 517
digestive gland cells: modulation of tyrosine kinase-mediated cell signaling. Gen. Comp.
518
Endocrinol. 122, 60-66. https://doi.org/10.1006/gcen.2001.7612 519
Canesi, L., Betti, M., Ciacci, C., Scarpato, A., Citterio, B., Pruzzo, C. and Gallo., G., 2002.
520
Signaling pathways involved in the physiological response of mussel hemocytes to bacterial 521
challenge: the role of stress-activated p38 MAP kinases. Dev. Comp. Immunol. 26, 325-334.
522
https://doi.org/10.1016/S0145-305X(01)00078-7 523
Canesi, L., Ciacci, C., Lorusso, L.C., Betti, M., Guarnieri, T., Tavolari, S. and Gallo, G., 524
2006. Immunomodulation by 17 beta-estradiol in bivalve hemocytes. Am. J. Physiol. Regul.
525
Integr. Comp. Physiol. 291, R664-673. https://doi.org/10.1152/ajpregu.00139.2006 526
Canesi, L., Ciacci, C., Vallotto, D., Gallo, G., Marcomini, A. and Pojana, G., 2010. In vitro 527
effects of suspensions of selected nanoparticles (C60 fullerene, TiO2, SiO2) on Mytilus 528
hemocytes. Aquat. Toxicol. 96 (2), 151-8. https://doi.org/10.1016/j.aquatox.2009.10.017 529
Châtel, A., Hamer, B., Talarmin, H., Dorange, G., Schroder, H.C. and Muller, W.E., 2009.
530
Activation of MAP kinase signaling pathway in the mussel Mytilus galloprovincialis as 531
biomarker of environmental pollution. Aquat. Toxicol. 96, 247-255.
532
https://doi.org/10.1016/j.aquatox.2009.11.002 533
Châtel, A., Talarmin, H., Hamer, B., Schroder, H.C., Muller, W.E. and Dorange, G., 2011a.
534
MAP kinase cell signaling pathway as biomarker of environmental pollution in the sponge 535
Suberites domuncula. Ecotoxicology. 20, 1727-1740. http://dx.doi.org/10.1007/s10646-011- 536
0706-1 537
Châtel, A., Hamer, B., Jaksic, Z., Vucelic, V., Talarmin, H., Dorange, G., Schroder, H.C. and 538
Muller, W.E. 2011b. Induction of apoptosis in mussel Mytilus galloprovincialis gills by 539
model cytotoxic agents. Ecotoxicology. 20, 2030-2041. https://doi.org/10.1007/s10646-011- 540
0746-6 541
Couleau, N., Techer, D., Pagnout, C., Jomini, S. , Foucaud, L. , Laval-Gilly, P. , Falla, J.
542
and Bennasroune, A., 2012. Hemocyte responses of Dreissena polymorpha following a short- 543
term in vivo exposure to titanium dioxide nanoparticles: preliminary investigations. Sci. Total 544
Environ. 438, 490-7. https://doi.org/10.1016/j.scitotenv.2012.08.095 545
Farré, M., Gajda-Schrantz, K., Kantiani, L. and Barceló, D., 2009. Ecotoxicity and analysis of 546
nanomaterials in the aquatic environment. Anal. Bioanal. Chem. 393,81-95.
547
https://doi.org/10.1007/s00216-008-2458-1 548
549
Hughes, R.N., 1969. A study of feeding in Scrobicularia plana. J. mar. biol. Ass. U.K. 49, 550
805-823. https://doi.org/10.1017/S0025315400037309 551
552
Jones, J. D., 1961. Aspects of Respiration in Planorbis Corneus L. And Lymnaea Stagnalis L.
553
(Gastropoda: Pulmonata). Comp. Biochem. Physiol. 4,1-29. https://doi.org/10.1016/0010- 554
406X(61)90042-1 555
556
Kyriakis, J.M. and Avruch, J., 2001. Mammalian mitogen-activated protein kinase signal 557
transduction pathways activated by stress and inflammation. Physiol. Rev. 81, 807-869.
558
https://doi.org/10.1152/physrev.2001.81.2.807 559
560
Kultz, D. and Avila, K., 2001. Mitogen-activated protein kinases are in vivo transducers of 561
osmosensory signals in fish gill cells. Comp. Biochem. Physiol. B. Biochem. Mol. Biol.
562
129,821-829. https://doi.org/10.1016/S1096-4959(01)00395-5 563
564
Mohd Omar, F., Abdul Aziz, H. and Stoll, S., 2014. Aggregation and disaggregation of ZnO 565
nanoparticles: Influence of pH and adsorption of Suwannee River humic acid. Sci. Total 566
Environ. 468–469,195–201. https://doi.org/10.1016/j.scitotenv.2013.08.044 567
568
Mouneyrac, C., Linot, S., Amiard, J.C., Amiard-Triquet, C., Métais, I., Durou, C., Minier, C.
569
and Pellerin, J., 2008. Biological indices, energy reserves, steroid hormones and sexual 570
maturity in the infaunal bivalve Scrobicularia plana from three sites differing by their level of 571
contamination. Gen. Comp. Endocrinol. 157(2), 133-141.
572
https://doi.org/10.1016/j.ygcen.2008.04.010 573
574
Peralta-Videa, J.R., Zhao, L., Lopez-Moreno, M.L., de la Rosa, G., Hong, J. and Gardea- 575
Torresdey, J.L., 2011. Nanomaterials and the environment: a review for the biennium 576
2008−2010. J. Hazard. Mater. 186, 1-15. https://doi.org/10.1016/j.jhazmat.2010.11.020 577
578
Rocha, M., Fernandes, C., Pereira, C., Rebelo, S.L.H., Pereira, M.F.R. and Freire, C., 2015.
579
Gold-supported magnetically recyclable nanocatalysts: a sustainable solution for the reduction 580
of 4-nitrophenol in water. RSC Adv. 5, 5131-5141. https://doi.org/10.1039/c4ra15865b 581
582
Roux, P.P. and Blenis, J., 2004. ERK and p38 MAPK-activated protein kinases: a family of 583
protein kinases with diverse biological functions. Microbiol. Mol. Biol. Rev. 68, 320-344.
584
https://doi.org/10.1128/MMBR.68.2.320-344.2004 585
586
Tella, M., Auffan, M., Brousset, L., Issartel, J., Kieffer, I., Pailles, C., Morel, E., Santaella, C., 587
Angeletti, B., Artells, E., Rose, J., Thiery, A., Bottero, J.Y., 2014. Transfer, Transformation 588
and Impacts of Ceria Nanomaterials in Aquatic Mesocosms Simulating a Pond Ecosystem.
589
Environmental Science & Technology. 48, 9004–9013. https://doi.org/10.1021/es501641b 590
591
Tella, M., Auffan, M., Brousset, L., Morel, E., Proux, O., Chaneac, C., Angeletti, B., Pailles, 592
C., Artells, E., Santaella, C., Rose, J., Thiery, A., Bottero, J.Y., 2015. Chronic Dosing of a 593
Simulated Pond Ecosystem in Indoor Aquatic Mesocosms: Fate and Transport of Ceo2 594
Nanoparticles. Environmental Science-Nano. 2, 653-663.
595
http://dx.doi.org/10.1039/C5EN00092K 596
597
Vale, G., Mehennaoui, K., Cambier, S., Libralato, G., Jomini, S. and Domingos, R.F., 2016.
598
Manufactured nanoparticles in the aquatic environment-biochemical responses on freshwater 599
organisms: A critical overview. Aquat. Toxicol. 170,162-174.
600
https://doi.org/10.1016/j.aquatox.2015.11.019 601
602 603 604 605 606 607
ACKNOWLEDGEMENTS 608
The research leading to these results has received funding from the European Research 609
Council under the European Union's FP7 Grant Agreement n.310584 (NANoREG project).
610
The authors wish to thank L. Izoret (ATILH, Paris, France) for providing cements, as well as 611
A. Guiot, S. Artous, S. Jacquinot and O. Sicardy (CEA, Grenoble, France) for their 612
participation to the characterization of the raw materials. The TEM used in this study was part 613
of Nano-ID platform which was funded by the EQUIPEX project ANR-10-EQPX-39-01. Part 614
of this study was also funded via the French ANR through the ANR-3-CESA-0014/
615
NANOSALT program, and the Excellence Initiative of Aix- Marseille University - 616
A*MIDEX, a French “Investissementsd'Avenir” program through its associated Labex 617
SERENADE project. This work was also a contribution to the OSU-InstitutPythéas. Finally, 618
the authors acknowledge the CNRS funding for the international research project IRP iNOVE.
619 620 621 622
SUPPORTING INFORMATION 623
624
Figure S1. Primary size and shape of TiO2MNMs determined by an ultra-high resolution 625
scanning electron microscopy (SEM) (A) and Transmission Electron Microscopy (TEM) (B).
626
Body-centered tetragonal anatase crystal structure of TiO2MNM observed by X-ray 627
diffraction (XRD) (C).
628 629
A B
Table 1. Concentrations of Ti, Ca, Al, Fe, Ca and Si after cement degradation in the total 630
(dissolved + particulate fractions) and soluble fraction solution 631
632
Ti Na Al Fe Ca Si
Total (mg.L-1) 89.91 1.89 54.06 12.10 2216.00 2407.80 Soluble fraction (<3 kDa)
(µg/l) 3.22 182.00 1.00 1.00 2.03E+05 3960.00 633
634 635
636
637
638
Figure S2. Evolution of the physical-chemical parameters in the water column of the marine 639
mesocosms. Temperature, Redox potential, dissolved oxygen, pH were measured during 640
phases I (stabilization) and II (contamination). Day 0 corresponds to the first dosing of NPs.
641
The grey surface is defined by the maximum and minimum values of each parameter, and the 642
dark line corresponds to the average values of the 9 mesocosms. One measurement was 643
performed every 5 min.
644 645
646
647
Figure S3. Evolution of the physical-chemical parameters in the water column of the 648
freshwater mesocosms. Oxidation-Reduction potential (ORP), dissolved oxygen, pH, and 649
conductivity were measured during phases I (stabilization) and II (contamination). Day 0 650
corresponds to the first dosing of NPs. The grey surface is defined by the maximum and 651
minimum values of each parameter, and the dark line corresponds to the average values of the 652
6 mesocosms. The conductivity were not merged since a different trend was observed in the 653
mesocosms exposed to cement leachate. One measurement was performed every 5 min.
654
655
656
Figure S4. Concentration of picoplankton in water column at 10 cm below water surface 657
(Top) and in surficial sediments(depth: 0.5 to 1 mm) (B)on a weekly basis.
658
Five mL of water and 15 mL of sediment were sampled, treated with formaldehyde (3.7%), 659
and stored at 4°C before counting. Before picoplankton counting, 1 mL of each water column 660
sample was centrifuged (5.9 × g at 4°C for 15 min), and 200 μL of each sediment sample was 661
treated with 800 μL of 0.1 mM sterile tetrasodium pyrophosphate and vortexed with a steel 662
ball for 30 seconds. For the counting, 10 μL of each sample was mixed with 5μL of 3μM 663
SYTO® 9 Green Fluorescent Nucleic Acid Stain and dropped on a glass slide. Concentration 664
of picoplankton was the mean ± standard deviation of five counts.
665 666 667 668